This repository contains the code and datasets for "MaxGIRF: Image Reconstruction Incorporating Concomitant Field and Gradient Impulse Response Function Effects", by Nam G. Lee, Rajiv Ramasawmy, Yongwan Lim, Adrienne E. Campbell-Washburn, and Krishna S. Nayak.
Nam Gyun Lee, University of Southern California, Dec 2021.
Example human and phantom datasets can be found on Zenodo:
-
Axial Cartesian spin-echo dataset
- meas_MID00273_FID03656_se_15b130_tra.h5
- noise_meas_MID00273_FID03656_se_15b130_tra.h5
-
Axial spiral spin-echo dataset
- meas_MID00275_FID03658_se_spiral_1102_ax_s24.h5
- noise_meas_MID00275_FID03658_se_spiral_1102_ax_s24.h5
-
Axial Cartesian GRE datasets for static off-resonance map estimation
- meas_MID00276_FID03659_gre_TE1.h5
- meas_MID00277_FID03660_gre_TE2.h5
- meas_MID00278_FID03661_gre_TE3.h5
- meas_MID00279_FID03662_gre_TE4.h5
- meas_MID00280_FID03663_gre_TE5.h5
- noise_meas_MID00276_FID03659_gre_TE1.h5
- noise_meas_MID00276_FID03660_gre_TE2.h5
- noise_meas_MID00276_FID03661_gre_TE3.h5
- noise_meas_MID00276_FID03662_gre_TE4.h5
- noise_meas_MID00276_FID03663_gre_TE5.h5
-
Sagittal Cartesian spin-echo dataset
- meas_MID00258_FID03641_se_15b130.h5
- noise_meas_MID00258_FID03641_se_15b130.h5
-
Sagittal spiral spin-echo dataset
- meas_MID00260_FID03643_se_spiral_1102_sag_s24.h5
- noise_meas_MID00260_FID03643_se_spiral_1102_sag_s24.h5
-
Sagittal Cartesian GRE datasets for static off-resonance map estimation
- meas_MID00261_FID03644_gre_TE1.h5
- meas_MID00262_FID03645_gre_TE2.h5
- meas_MID00263_FID03646_gre_TE3.h5
- meas_MID00264_FID03647_gre_TE4.h5
- meas_MID00265_FID03648_gre_TE5.h5
- noise_meas_MID00261_FID03644_gre_TE1.h5
- noise_meas_MID00261_FID03645_gre_TE2.h5
- noise_meas_MID00261_FID03646_gre_TE3.h5
- noise_meas_MID00261_FID03647_gre_TE4.h5
- noise_meas_MID00261_FID03648_gre_TE5.h5
-
Mat files containing a static off-resonance map obtained with NLINV estimation
- B0map_nlinv_min1.0e-06_axial.mat
- B0map_nlinv_min1.0e-06_sagittal_ro0.mat (ro := remove oversampling: 1=yes, 0=no)
- B0map_nlinv_min1.0e-06_sagittal_ro1.mat
Once datasets are downloaded, organize datasets in this structure:
|---path-to-dataset (e.g., D:\lowfield\NHLBI\data\20201102_NV_brain)
|---meas_MID00260_FID03643_se_spiral_1102_sag_s24.h5
|---meas_MID00275_FID03658_se_spiral_1102_ax_s24.h5
|---noise_meas_MID00260_FID03643_se_spiral_1102_sag_s24.h5
|---noise_meas_MID00275_FID03658_se_spiral_1102_ax_s24.h5
|---B0map_nlinv_min1.0e-06_axial.mat
|---B0map_nlinv_min1.0e-06_sagittal_ro0.mat
|---B0map_nlinv_min1.0e-06_sagittal_ro1.mat
Update variables src_directory
and mida_directory
in figure3\demo_2d_spiral_simulation.m
.
src_directory = 'path-to-this-package';
mida_directory = 'path-to-this-package\thirdparty\MIDA_v1.0\MIDA_v1_voxels';
Run figure3\demo_batch_2d_spiral_simulation.m
Run figure3\demo_generate_figure3.m
to generate Figure 3.
Update variables in demo_non_cartesian_recon_human_axial.m
.
src_directory = 'path-to-this-package';
ismrmrd_directory = 'path-to-ISMRMRD-package';
ismrmrd_noise_fullpath = 'path-to-dataset\noise_meas_MID00275_FID03658_se_spiral_1102_ax_s24.h5';
ismrmrd_data_fullpath = 'path-to-dataset\meas_MID00275_FID03658_se_spiral_1102_ax_s24.h5';
siemens_dat_fullpath = 'path-to-dataset\meas_MID00275_FID03658_se_spiral_1102_ax_s24.dat';
B0map_fullpath = 'path-to-dataset\B0map_nlinv_min1.0e-06_axial.mat';
Run demo_non_cartesian_recon_human_axial.m
.
Run demo_cartesian_recon_SE_axial.m
.
Update variables in demo_non_cartesian_recon_human_sagittal.m
.
src_directory = 'path-to-this-package';
ismrmrd_directory = 'path-to-ISMRMRD-package';
ismrmrd_noise_fullpath = 'path-to-dataset\noise_meas_MID00260_FID03643_se_spiral_1102_sag_s24.h5';
ismrmrd_data_fullpath = 'path-to-dataset\meas_MID00260_FID03643_se_spiral_1102_sag_s24.h5';
siemens_dat_fullpath = 'path-to-dataset\meas_MID00260_FID03643_se_spiral_1102_sag_s24.dat';
B0map_fullpath = 'path-to-dataset\B0map_nlinv_min1.0e-06_sagittal_ro0.mat';
Run demo_non_cartesian_recon_human_sagittal.m
.
Run demo_cartesian_recon_SE_sagittal.m
.
Run figures7&8\demo_generate_figure7.m
and figures7&8\demo_generate_figure8.m
to generate Figures 7 and 8, respectively.